Hi all,

Looking for suggestions. Both my Gt-31's seem to be recording very rough tracks. This has been happening since a software up date. Possibly settings issue?!?! Not recording fast enough?!?!

What ver? Can you roll back one GPs & then compare ?

Roll back one gps? What does that mean
Both gps units have come up with very similar image, varying numbers for alphas

On the RHS of your screen grab, is a lot of selected (blue) rows. On that table, right-click and enable the "time" column... then post another screen grab showing the timestamps.

There are up to 5 sec gaps between points. What would cause this?

check the interval, should be 1 sec (not 5)

Same alpha, different GPS

pretty sure it looks like its only logging a posi every 5 seconad in the other one, i.e. the interval is set to 5 instead of 1. unfortunately the setup of the one GPS is going to invalidate ALL your cranking alpha results.

Yep stringer. Already pulled the PB back to a zero!! Any doubts from my point of view deserve a zero. If I did it once I'll do it again, if not it's data. Happy days!!

Still looking for answers as to why this happened though.

if you were wearing two GPS's and had one setup right then you shouldnt have to do that. i think under the data logger screen is the interval setting which should be set to 1sec (not 5sec)

They were both set on 1. The only thing was the minimum speed was on 5 (which I have now changed), but should not have made a difference as I had no reading that low.

Any other setting changes that might stop this happening would be appreciated.

Tomski, make sure that you have 1sec in both datalogger and memory card. (sorry if you have already done this. Have you done a save to card from the datalogger and compared the tracks?
Is Log switch set to On or On-fix?

Both gps were on on-fix. After reading sailquicks advice in another thread am running both on just 'ON'.

What will saving to card do Lao?? I will try it but just trying to get a better understanding

You have two sets of data.
1 is saved directly to the card and is a single file for that session.
2 is the datalogger (internal memory) which stores data for about the last 4.5 hrs over how ever many sessions unless you have cleared it.
If everything is good they should be the same data if you have the same settings (intervals, min speed etc)
If you save to card you transfer the datalogger to a single file on the card. People normally do this if the card has not recorded for some reason. Uploading this file to KA72 for example would show multiple sessions.


Might just pick up any irregularities in settings?

Must have clear data logger when I set them up again, but thanks for the advice. I'm hoping with the new settings next files should be sweet. Hopefully 25 knot alpha.

Thanks for the help Lao.

where were you wearing them Tom? Looks like one gps had a better satellite view than the other. "On fix" will give those results if the GPS looses satellites.

Select to expand quote

decrepit said...
where were you wearing them Tom? Looks like one gps had a better satellite view than the other. "On fix" will give those results if the GPS looses satellites.

There are other columns of data that can be shown - in particular "sats"... when the value drops below 5, you might still have a *lock* but the data-point will be quite inaccurate (and should be dropped) Another column is "HDOP" - larger is worse. "secs" can be useful if the value is something not close to 1.0.

Sailed again today with both setup right. Still the tracks look very jaggered (compared to mates file).

Mathew, I'm getting 6-9 sat. My worse HDOP is 1.1. and 1 sec all the time. I uplaoded both files to ka72 and got precise results (same as realspeed) for everything but alphas.
Why is my realspeed showing really hooky, short alphas now?

I think I've got it as good as it gets???

Mike,

I'm wearing it on my arm, as I always have.
Whats your take on the alphas? At what point do you not include an alpha? Greater then 1 sec between point? Spikes are obvious, but is what I have a spike?

Select to expand quote

Tomasz80 said...
Mike,

I'm wearing it on my arm, as I always have.
Whats your take on the alphas? At what point do you not include an alpha? Greater then 1 sec between point? Spikes are obvious, but is what I have a spike?

Hi Tom, If you would like to email me the files (SBN and SBP) I will try to figure out what is happening and let you know.

By the way, what is the version # of the updated firmware?

sailquik(at)hotmail(dot) com

Thanks Daffy. I have V1.4

Will send the tracks to you very soon

Tom I just downloaded both your files off KA72.
The first thing I noticed was that your trackpoint speed measurement was all over the place. The second thing I noticed was that your GPS point recording interval was up and down by a fraction of a second which will be causing this. I took this screenshot in GPS Results which I think displayed it better. The second is one of my files showing how mine reads.



The electronics geek in me would say that the clock circuit in your GPS is breaking down but because both your files are doing this I think it must be Firmware related

Select to expand quote

BundyBear said...
The electronics geek in me would say that the clock circuit in your GPS is breaking down but because both your files are doing this I think it must be Firmware related

Neither - and the clock rate wont matter... in fact if it always was less than 1.0, then the data gets *more accurate*. The variation in sats-count will be the cause, ie: the GPS algorithms work by having a "smooth" data feed -> loosing a satellite isn't really a smooth thing.

It might be worthwhile taking the GPS's for a walk, turn some corners, etc. then look at the data -> might be a simple "mounting" problem, or some water interference (eg: GPS under a wetsuit).

Thomas,
Below are two snapshots of Realspeed for similar open water sessions from myself and Kel (FNQ Mob). I run V1.3(B1123T) and Kel runs V1.4(B0803T) firmware. I have deliberately kept the older version as I reckon the V1.4 gives some unusual readings similar to what you've been experiencing (tracks are very spikey for the V1.4 firmware in comparison to V1.3) Don't really know if this is an issue and maybe Matthew or Daffy could elaborate or explain difference between firmware versions. It'd be great to have two GPS's set up with different firmware versions for direct comparison.

V1.3 firmware session on realspeed snapshot (Grid fields for both: Grid/Zoom/Speed/Doppler/Sats/Smoothed Speed/Nodes)


V1.4 firmware Realspeed Snapshot

Thanks for the effort Adam. I'm starting to think its a combo of 2 gps units in a different case that slides to the underside of my arm.[}:)]. Are they that sensative. I use an aqua pack and wear it on the top/outside of my arm!?!

Mathew, I don't cover them with anything, just the aqua pack, so inteference can only come from my body or the case?!?!.

I will try the walk around test tonight to see.

Thanks Sausage. I would like to see this explained (diff between 1.3 &1.4). Your track seems very similar to my friends file which I'm comparing to. I will check what version he is running.

Select to expand quote

Tomasz80 said...
Thanks Sausage. I would like to see this explained (diff between 1.3 &1.4). Your track seems very similar to my friends file which I'm comparing to. I will check what version he is running.

See if you can borrow his GPS with the V1.3 firmware and take for a sail with your GPS (V1.4) and then we'd have some definitive data to compare.

BTW - I wear GT-31 on my upper arm and never get spikes or loss of data.

Hi Tom. I got your files.
I have not see the less than exactly 1 second interval for a very long time. We used to see that occasionally but I think it was with the GT-11 or the first GT-31 firmware.
I think this is the indicator of what the problem is but can't figure it out yet.
I can't see anything else obvious apart from the fact that the SDOP (+/- in GPS-Results) goes way too high during all your alphas. This despite there being an adequate number of satellites in contact. It is not unusual to see a slight increase on SDOP during the gybe in Alphas, but these numbers are much higher than usual.
The first file you sent me would not even register results in GPS-Results (the one you labeled 'Alpha ****ups') It will open on RealSpeed, but of course RealSpeed can't display the SDOP data - (I think this was something Mal intended to build in but has never got around too).

If I were you I would try updating the firmware again paying particular attention to the procedures in the 'how to' instructions on Tom Chalko's website (mtbest.net). It may be that something went wrong during the first update.

Select to expand quote

sausage said...
Thomas,
Below are two snapshots of Realspeed for similar open water sessions from myself and Kel (FNQ Mob). I run V1.3(B1123T) and Kel runs V1.4(B0803T) firmware. I have deliberately kept the older version as I reckon the V1.4 gives some unusual readings similar to what you've been experiencing (tracks are very spikey for the V1.4 firmware in comparison to V1.3) Don't really know if this is an issue and maybe Matthew or Daffy could elaborate or explain difference between firmware versions. It'd be great to have two GPS's set up with different firmware versions for direct comparison.

Snags, those pics are quite telling... v1.3 shows 9 sats consistently, while v1.4 shows less sats AND it losing lock.

EDIT: That said, these tracks are for different days, so you *cannot* do any type of serious comparison.

Daffy, any chance you might be able to downgrade to do some testing?

Select to expand quote

mathew said...

sausage said...
Thomas,
Below are two snapshots of Realspeed for similar open water sessions from myself and Kel (FNQ Mob). I run V1.3(B1123T) and Kel runs V1.4(B0803T) firmware. I have deliberately kept the older version as I reckon the V1.4 gives some unusual readings similar to what you've been experiencing (tracks are very spikey for the V1.4 firmware in comparison to V1.3) Don't really know if this is an issue and maybe Matthew or Daffy could elaborate or explain difference between firmware versions. It'd be great to have two GPS's set up with different firmware versions for direct comparison.

Snags, those pics are quite telling... v1.3 shows 9 sats consistently, while v1.4 shows less sats AND it losing lock.

Daffy, any chance you might be able to downgrade to do some testing?

Mathew,
Apologies for not clarifying but these two sessions were on different days and spots so sat no.s not really comparable. I was more querying the jaggered tracks and irregular 1 second recordings with the V1.4 firmware against the smoother V1.3 one.

Hi Sausage. That v1.4 data does look remarkably similar to Tom's. Have you seen any other v1.4 GPS that give such data?

I just compared some of my v1.3 v's v1.4 data from the same session and I don't see those sorts of results. They are almost identical and the same as comparing two v1.4 GPS. (At the moment I am using two GT-31's with v1.4 and one with v1.3 - I have two in my helmet and one in a bag on my arm to look at.)

The only difference I see is that the two in the helmet always give smoother tracks than the one on the arm, especially during Alphas, were the arm mounted GPS often gets SDOP speed error range much higher during the gybe. This is irrespective of the firmware mix.

I imagine 95% of users have their GPS mounted in an arm bag which makes me think that an awful lot of Alphas would have SDOP figures over 1.5. This is certainly getting into the 'suspect' range and anything over 3 is definitely untrustworthy.

Unfortunately, RealSpeed does not have this filter as far as I can tell......

Here are results an Alpha, just a normal gybe really, at Lake George a couple of weeks ago. GPS N1 and N2 were in the helmet. AD3 was on the arm.







Explanation:
The first two show a smooth track with little change during the gybe in the SDOP (+/- column) and dotted line each side of the speed graph.
The third, which was on the arm, shows a jump to over 3Kts SDOP probable error at the gybe and you can see this in the speed graph with the dotted lines a long way apart. Note also that since I have my SDOP filter set to 1.5, GPS-Results tells me this is an error/invalid run by not drawing the speed graph over those points.

Speculation: I think the difference is because the arm makes a much more abrupt change of direction during the gybe than the GPS in the helmet. This affects the calculation of the HDOP and therefore the SDOP. Mathew can can explain it much better than me as he was the one who pointed it out to me.